TopxAI

Loading...

Docs · Models

Kimi K3.1 Pro

Call kimi-k3.1-pro on TopxAI: model migration, fixed token prices, always-on reasoning, output caps and image-input limits.

Use kimi-k3.1-pro with a kimi-official or Auto-route key. This model replaces kimi-k3; the old ID returns 403. A key restricted to the old ID needs an explicit allowlist change. Saved client configurations must also be updated.

Send a text request

curl https://ai.topxea.com/v1/chat/completions \
  -H "Authorization: Bearer $TOPXAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kimi-k3.1-pro",
    "reasoning_effort": "low",
    "max_completion_tokens": 4096,
    "messages": [{"role":"user","content":"Explain the difference between a process and a thread."}]
  }'

Chat Completions is the direct text interface. Anthropic-compatible clients can use /v1/messages; TopxAI converts that request to Chat Completions. Preserve complete assistant messages, including reasoning and tool-call fields, when constructing subsequent turns.

Responses requests

POST /v1/responses accepts kimi-k3.1-pro with the full conversation supplied in input. Set store: false and use stateless requests; server-side conversation persistence and a complete Codex workflow have not been verified. Set reasoning through reasoning: {"effort":"low"} when using this format.

Limits and input formats

The context window is 1,048,576 tokens. TopxAI publishes a 131,072-token output limit, also the provider's default, to retain prompt room. The provider documents a 1,048,576-token hard ceiling within the combined context budget; that is not TopxAI's published output limit. Reasoning is always on, with low, high and max as the supported efforts and max as the default. Omit sampling controls rather than carrying values from a different model. These limits follow the official model guide, checked October 24, 2026.

For image input, use a Chat Completions content array containing an image_url part with a base64 data URI. Studio's image attachment button produces that format. Public HTTP image URLs are not a supported Kimi input. The provider also documents ms:// file references and video input, but those depend on files uploaded to the provider account; this page does not offer a TopxAI file-upload workflow for them. See Kimi's vision documentation.

Fixed prices and cache writes

Default USD prices per million tokens are $2.40 input, $0.24 cache read, $2.40 cache write and $12 output. These retain TopxAI's previous selling rates and equal 120% of the new model's official reference, or 20% higher. There is no long-context pricing tier.

Only the default 5-minute cache-write category is offered here. TopxAI does not forward a request for the provider's longer TTL. Cache hits and writes are charged only as reported by upstream usage. Kimi's caching guide describes the provider's separate TTL categories; the TopxAI pricing page shows the categories sold on this route.

Models: kimi-k3.1-pro · Models & pricing

Related

Previous: deepseek-flash-Abliterated · Next: Top up with a card

Other languages: 简体中文